Pakistan: Ex Post Assessment of Longer-Term Program Engagement

This 2005 Article IV Consultation for Pakistan reports that macroeconomic performance continues to be strong, but inflation and external pressures are evident. High growth, including in agriculture, is estimated to have raised disposable incomes. The authorities’ ambitious growth objectives require substantial increases in investment, including from the private sector. Public investment has been increased to ease bottlenecks in transportation, irrigation, and energy. In the short term, macroeconomic policies need to be tightened somewhat to counter inflation and external pressures.

Abstract

This 2005 Article IV Consultation for Pakistan reports that macroeconomic performance continues to be strong, but inflation and external pressures are evident. High growth, including in agriculture, is estimated to have raised disposable incomes. The authorities’ ambitious growth objectives require substantial increases in investment, including from the private sector. Public investment has been increased to ease bottlenecks in transportation, irrigation, and energy. In the short term, macroeconomic policies need to be tightened somewhat to counter inflation and external pressures.

I. Introduction and Overview

1. Pakistan has been one of the most prolonged users of Fund resources. In the last 46 years, the country has been under Fund-supported programs for about half of the time (Figure 1). Before 1988, Fund-supported programs were infrequent and mostly short term in nature. But as growth faltered in the late 1980s, and the failure to redress fiscal and external imbalances resulted in an unsustainable debt burden, Pakistan became more heavily dependent on Fund support. The period 1988–99 was marked by regional problems and political instability—there were as many Fund programs as governments (eight each). Economic growth slowed, poverty increased, and governance problems became increasingly evident. However, from 1999, the new government of General Pervez Musharraf began addressing the economic difficulties with a new commitment to turn around past trends. The economy started to recover and progress has been made in structural reforms, including in building institutions.

Figure 1.
Figure 1.

Pakistan: Real Per Capita GDP Growth and Fund Programs, 1960–2004

Citation: IMF Staff Country Reports 2005, 409; 10.5089/9781451830682.002.A003

Sources: Pakistani authorities; and IMF.

2. All but the last two Fund-supported programs suffered from substantial policy slippages. Of the SDR 3.6 billion committed during 1988–99, 55 percent was not drawn as programs went off-track soon after their start (Table 1). Programs were either extended to accommodate delays in the completion of reviews, or were ended prematurely, to be replaced by new arrangements. In contrast, both the 2000 Stand-by Arrangement (SBA) and the 2001 three-year arrangement under the Poverty Reduction and Growth Facility (PRGF) were successfully completed without any extensions. All disbursements were made available upon completion of the program reviews, but the authorities decided not to draw the last tranche under the PRGF arrangement or seek a successor arrangement in light of Pakistan’s improved external situation (Figures 2 and 3). The improved economic situation enabled Pakistan to re-enter the international financial markets in 2004 to help cover its financing needs, in addition to ongoing support from the World Bank and the Asian Development Bank (AsDB) and a few bilateral donors.

Figure 2.
Figure 2.

Pakistan: Fund Financial Support, 1984–2004

(In millions of SDR)

Citation: IMF Staff Country Reports 2005, 409; 10.5089/9781451830682.002.A003

Source: IMF.
Figure 3.
Figure 3.

Pakistan: Outstanding Loans and Credits, 1984–2004

(In millions of SDR)

Citation: IMF Staff Country Reports 2005, 409; 10.5089/9781451830682.002.A003

Source: IMF.
Table 1.

Pakistan: History of Lending Arrangements, 1958–2004

article image
Source: IMF.

Blended arrangement.

Total number of years under arrangement, with blended arrangements counted as a single arrangement.

3. The Independent Evaluation Office (IEO), in its 2002 report evaluating the prolonged use of Fund resources (refer to www.imf.org, Vol. II), assessed Pakistan’s program performance until 1999. It found that the limited achievements of successive programs before 1999 were associated with considerable political and regional instability, but also reflected design and implementation problems. A basic problem was that successive governments did not have sufficient ownership of the programs and were often unable to sustain the reform and adjustment effort for more than a short period of time. In addition, programs typically suffered from overoptimistic macroeconomic assumptions and proved to be unrealistic about the pace and breadth of the structural reform agenda that was likely to be implemented by the government. Notably, limitations in implementation capacity were not sufficiently taken into account in program design, and the reform agenda was not prioritized enough.

4. According to the IEO, most of the design and implementation problems had their roots in institutional factors. In particular, decisions regarding the Fund’s involvement in Pakistan appeared to have been heavily influenced by (a) geopolitical considerations; (b) the difficulty of not giving a new government the initial benefit of the doubt as regards its stated policy objectives; (c) a reluctance to disengage when a country faces difficulties; (d) the presumption of a short- to medium-term involvement only; and (e) collaboration with the World Bank that did not always produce an effective integration of priorities and timeframes in dealing with key structural issues. More generally, the IEO concluded that there had been too few attempts to “step back” and reconsider the overall strategy on the basis of a candid assessment of previous programs. The IEO considered the main lesson from Pakistan’s prolonged-use experience to be a need for greater selectivity, based on a candid assessment of ownership, and disengaging from a country, if necessary. Also, greater efforts were needed to tailor program design to the specific circumstances and long-term needs of a country, focusing on key institutional changes.

5. This report mainly focuses on the post-1999 period, where the IEO left off. After a brief description of conditionality in Fund-supported programs during 2000–04 (Section II.A), economic achievements as well as those areas where progress has been limited or absent are discussed (Section II.B). The main issues addressed by the Ex Post Assessment are covered in Section III, starting with what caused the change in performance after 1999 (Section III.A). Section III.B asks whether Fund conditionality was appropriate and Section III.C focuses on the design of the monetary program. Other issues discussed are whether Fund-recommended policies helped to reduce poverty (Section III.D); was the size of Fund financial support appropriate (Section III.E); did Fund-Bank collaboration contribute to the success of the program (Section III.F); and whether the Fund’s involvement in Pakistan changed following the 2002 IEO report (Section III.G). Section IV concludes with lessons for the Fund that can be derived from Pakistan’s experience.

II. The Facts

A. Program Conditionality

6. Conditionality in Pakistan’s Fund-supported programs became increasingly extensive after 1999. In the 1990s, programs included fairly standard quantitative conditionality: a floor on the net foreign assets and ceilings on the net domestic assets of the State Bank of Pakistan (SBP), the overall budget balance, government bank borrowing, and net banking sector claims on public sector enterprises, as well as on government external borrowing and the accumulation of external arrears. To address shortcomings in performance in key areas, quantitative conditionality was augmented in the 2000 SBA and 2001 PRGF by a floor on Central Board of Revenues (CBR) revenues and a ceiling on the accumulation of budgetary arrears to the Water and Power Development Authority (WAPDA); and—in the PRGF—limits on SBP’s foreign exchange reserves held with banks and on foreign currency swap and forward sales between the SBP and residents.2 Nevertheless, the amount of quantitative conditionality was not out of line compared with the average PRGF program.

7. In the 1990s, structural conditionality had typically been set in the form of broad commitments in the letter of intent or structural benchmarks. But as ownership and implementation was often weak, the response of the staff was to gradually try and close as many loopholes as possible, through an increase in the number of both macroeconomic and structural conditions, especially through much larger recourse to prior actions and performance criteria (PCs).

8. structural conditionality peaked in the 2000 SBA. The total of 56 PCs, benchmarks and prior actions in this arrangement was nearly three times the Fund-wide average of 19 conditions in GRA-supported programs (Table 2 and Figure 4).3 The large number of conditions was justified on grounds of a need to establish a strong track record by making up for past policy slippages within a relatively short time frame (see below).

Figure 4.
Figure 4.

Pakistan: Structural Conditions per Year, 2000–04

Citation: IMF Staff Country Reports 2005, 409; 10.5089/9781451830682.002.A003

Source: IMF.
Table 2.

Pakistan: Structural Conditionality, 2000–04 1/

article image
Source: IMF.

Number of conditions effective per program year. Continuous performance criteria in effect in more than one year are counted as one in each year they are in effect. Prior actions that make up for missed performance criteria or benchmarks are not counted.

Average of GRA-supported programs in first column (2001) and average of PRGF-supported programs in rest of table (2002–04).

9. Following the 2000 SBA, structural conditionality was reduced gradually because of Pakistan’s improved performance, as well as rising doubts in the Fund about the effectiveness of large policy matrices of conditions. The new government generally delivered on its commitments and the number of structural conditions was steadily reduced during the 2001–04 PRGF arrangement. This steady decline in the number of conditions also reflected the overall drive in the Fund to streamline conditionality. Nevertheless, only in the last program year did the number of structural conditions (15) approach the Fundwide average for PRGF-supported programs (14).

10. Structural conditionality became increasingly focused. Measures to improve the tax system dominated conditionality in the 2000 SBA and the 2001 PRGF arrangements, reflecting the critical importance attached to increasing the revenue intake (Figure 5). Measures falling in the category of general economic management—such as the promulgation of a budget consistent with program understandings or a depreciation of the exchange rate—were numerous in the 2000 SBA, but declined sharply as macroeconomic performance strengthened. During the PRGF arrangement, conditionality focused also on financial sector reforms and public enterprise reforms. While efficiency factors were considered, ultimately these measures aimed to achieve budget savings, by reducing the financing costs of the government by aligning yields on National Savings Scheme (NSS) instruments to market rates, and by restructuring the energy sector, thus reducing its reliance on budget support.

Figure 5.
Figure 5.

Pakistan: Composition of Structural Conditions, 2000–04

Citation: IMF Staff Country Reports 2005, 409; 10.5089/9781451830682.002.A003

Source: IMF.

11. Program implementation in 2001–04 compared favorably to the average GRA-and PRGF-supported programs. However, no review of the 2000 SBA or the 2001–04 PRGF arrangement was completed without requesting at least one waiver. On average, two waivers were requested per review. Still, waiver rates, on average, were comparable to, or even lower than, those in average Fund-supported programs (Table 3 and Figure 6).4 In part, the frequency of waiver requests was a reflection of the large number of conditions. Moreover, waivers were often needed because of delays in implementation rather than an outright lack of implementation. This was particularly true for energy sector measures, which typically required more time to overcome political or technical hurdles, causing an increase in the waiver rate for structural conditions in the last year of the PRGF arrangement. Looking at the share of lapsed measures—measures that never got implemented—implementation in Pakistan compared favorably relative to the average Fund-supported program during this period: only 6 percent of structural performance criteria were not implemented at all, compared to a Fund-wide average of 14–15 percent. Compliance with quantitative macroeconomic conditionality was also, on average, better than the average for Fund-supported programs.

Figure 6.
Figure 6.

Pakistan: Implementation of Conditionally, 2000–04

Citation: IMF Staff Country Reports 2005, 409; 10.5089/9781451830682.002.A003

Source: IMF.
Table 3.

Pakistan: Implementation of Conditionality, 2000–04

article image
Source: IMF.

B. Goals and Achievements

Program objectives

12. By the late 1990s, Pakistan’s economy had run aground. Per capita GDP growth had slowed to below 1 percent on average during 1996/97–1998/99. By end-1998/99, the country’s foreign exchange reserves were virtually exhausted and covered only four weeks of imports. Public debt had risen to about 90 percent of GDP and debt service obligations could not all be met in 1998 when Pakistan came under sanctions following its nuclear tests. Using the official poverty line as a yardstick, the headcount poverty rate increased from 25 percent in the early 1990s to 32 percent in 2000/01.

13. The new government that assumed office in 1999 aimed to make a clean break from past weak and erratic policy implementation. In late 1999, the government formulated and started to implement policies aimed at reducing Pakistan’s high debt burden to achieve sustainability and create more room for social spending. The government also requested support from the Fund and a SBA was approved in November 2000. The immediate objective of the SBA was to reduce quickly macroeconomic imbalances through a tightening of both fiscal and monetary policies. The exchange rate was to be allowed to adjust to market forces to help strengthen the external position. The authorities were also to press ahead with structural reforms to enhance the economy’s growth potential and reduce poverty. As it was recognized that addressing Pakistan’s deep structural problems required a long-term effort, the 2000 SBA was intended to establish a track record and pave the way for a medium-term program that could be supported under the PRGF.

14. Pakistan’s request for a PRGF arrangement was approved in December 2001. The PRGF arrangement had three main objectives: increasing growth, reducing vulnerability to shocks, and improving social indicators. Accordingly, macroeconomic policies aimed at further fiscal adjustment to reduce the public debt overhang, while simultaneously increasing the revenue intake and reducing transfers to state-owned entities to create additional room for social spending. External support, including a Paris Club rescheduling, was obtained to alleviate Pakistan’s external debt burden. Monetary policy aimed at low inflation, with a view to replenishing foreign exchange reserves and thus reducing external vulnerabilities. Structural reforms were aimed at improving investor confidence and creating an environment in which the private sector could flourish.

What has been achieved?

15. Economic growth has recovered convincingly from the crisis of 1998. The rate of real per capita GDP growth accelerated from close to zero in 2000/01 to over 6 percent in 2004/05 (Table 4 and Figure 7). Initially, growth was driven by a strong rebound in exports, which recorded double-digit growth rates, and virtually closed the gap between exports and imports of goods and nonfactor services. Reflecting this, manufacturing recorded high growth rates. In later years, favorable weather conditions also contributed to a strong pick up in agricultural production adding to the overall strong growth performance. Although the investment ratio has remained broadly unchanged at 16–17 percent of GDP since 1999/2000, the contribution of total factor productivity to growth has increased sharply. Rising real incomes eventually contributed to a strong pick up in private consumption and in 2004/05 domestic demand took over as the main driving force of growth. With the economy heating up, imports surged—showing a sharp increase in shipments of consumer goods, as well as of machinery—resulting in a shift of the current account balance (excluding official transfers) from a surplus of almost 4 percent of GDP in 2002/03 to a deficit of almost 2 percent of GDP in 2004/05 (Figure 8).

Figure 7.
Figure 7.

Pakistan: Economic Growth, 1999/2000–2004/05

Citation: IMF Staff Country Reports 2005, 409; 10.5089/9781451830682.002.A003

Source: Pakistani authorities.
Figure 8.
Figure 8.

Pakistan: Balance of Payments and International Reserves, 1999/2000–2004/05

Citation: IMF Staff Country Reports 2005, 409; 10.5089/9781451830682.002.A003

Source: Pakistani authorities.
Table 4.

Pakistan: Selected Economic Indicators, 1999/2000–2004/05

article image
Sources: Data provided by the Pakistani authorities; and Fund staff estimates.

Program projections are expressed in percent of the rebased GDP. For this, the actual program figures are multiplied by the ratio of actual GDP before rebasing to actual GDP after rebasing.

IMF Country Report No. 01/24.

IMF Country Report No. 01/222.

16. The fiscal position was strengthened to reduce indebtedness and to create additional room for social spending. Fiscal consolidation was achieved to a large extent by increases in nontax revenues and savings on the interest bill. The latter reflected the restructuring of external debts, falling domestic real interest rates, and the decline in the government debt ratio from almost 90 percent of GDP in 2000/01 to 60 percent of GDP in 2004/05. The overall fiscal deficit (excluding grants) declined from 6.6 percent of GDP in 1999/2000 to about 3 percent of GDP in 2004/05 (Figure 9). Despite this adjustment, social- and poverty-related expenditures were raised by over 1 percentage point of GDP over the last three years, or by over 30 percent in real per capita terms. Nevertheless, the level of social spending remains quite low in comparison with other developing countries. While the goals that had been set at the start of the 2001 PRGF-supported program for raising social spending and reducing the debt burden were realized, more could have been accomplished if greater progress had been made in raising the tax-to-GDP ratio or reducing transfers to state-owned enterprises (see below).

Figure 9.
Figure 9.

Pakistan: General Government Budget, 1999/2000–2004/05

Citation: IMF Staff Country Reports 2005, 409; 10.5089/9781451830682.002.A003

Source: Pakistani authorities.

17. External vulnerabilities have been greatly reduced. Fiscal consolidation and debt relief granted by external creditors resulted in a decline in the ratio of external public and publicly guaranteed debt to GDP to 28 percent by end-2004/05, down from the peak of 45 percent three years earlier (Figure 10). The balance of payments has turned around with strong export growth and sizable remittances, and international reserves were rebuilt to cover about five months of imports of goods and nonfactor services and over 250 percent of short-term external liabilities (up from a mere 11 percent in 1999/2000) at end-2004/05.

Figure 10.
Figure 10.

Pakistan: Debt Indicators, 1999/2000–2004/05

(In percent of GDP)

Citation: IMF Staff Country Reports 2005, 409; 10.5089/9781451830682.002.A003

Source: Pakistani authorities.

18. Inflation remained low and stable until mid-2003, but has become a concern more recently. Initially, slack in the economy, that reflected the slow growth experienced in the late 1990s, and a steady increase in money demand, as confidence recovered, allowed the SBP to reduce interest rates, while inflation was kept in the low single digits (Figure 11). In the second half of 2003, however, with the economy starting to grow strongly, inflation began to pick up (from a low level). The SBP was slow to respond by tightening monetary policy, mindful not to choke off the economic recovery. As a result, key policy interest rates have been negative in real terms since 2003. The 12-month rate of inflation reached double digits in March 2005 for the first time since September 1997 and is currently at 8.4 percent.

Figure 11.
Figure 11.

Pakistan: Inflation and Interest Rates, Jul 1999–Aug. 2005

Citation: IMF Staff Country Reports 2005, 409; 10.5089/9781451830682.002.A003

Source: Pakistani authorities.

19. Good progress was made in some areas of structural reform. In the years prior to 1999, structural reforms had been started in the areas of interest rate, trade, and capital account liberalization. After 1999, the authorities continued to move ahead with liberalizing Pakistan’s trade and exchange regime, and strong progress was made in financial sector reform. The banking sector is now largely privately owned and has become more competitive and efficient, following restructuring and privatization. This was complemented by a strengthening of supervision and prudential regulations, as noted in the recent Financial System Stability Assessment (FSSA; IMF Country Report No. 04/215). With reforms and improving macroeconomic performance, banks’ financial soundness indicators improved and dollarization was reduced substantially. The tax and customs administration has been strengthened and made more taxpayer-friendly, while customs tariffs and tax rates were reduced and rationalized. The Fiscal Responsibility Law, which established legal targets for reducing the public debt burden and for reporting requirements, was approved in 2005. In addition, the role of the state in the economy has been reduced and governance improved through strengthening macroeconomic institutions, an extensive privatization program, and the establishment of improved regulatory frameworks in such sectors as telecommunications, electricity, and oil and gas.

What has not been achieved?

20. A major disappointment has been the lack of progress in raising tax revenues as a share of GDP. One of the objectives of the last two programs was a significant expansion of the tax base to create room for achieving the twin goals of higher social spending and debt reduction. However, the tax-to-GDP ratio has remained broadly unchanged at about 11 percent since the late 1990s and dipped in 2004/05 (Figure 12). Revenue reforms may have resulted in a more efficient system of taxation of existing taxpayers and more modern tax administration, but the number of taxpayers has not increased enough to achieve an increase in the tax-to-GDP ratio. Large parts of the agricultural and services sector still remain outside the tax net, reflecting political difficulties in taxing these sectors (see below).

Figure 12.
Figure 12.

Pakistan: Tax Revenues, 1993/94–2004/05

(In percent of GDP)

Citation: IMF Staff Country Reports 2005, 409; 10.5089/9781451830682.002.A003

Source: Pakistani authorities.

21. Progress was also disappointing in addressing the quasi-fiscal deficit of public utilities. The largest recipients of budget support by far are the companies in the power sector. The sum of subsidies and net lending to these companies has remained well over 1 percent of GDP. Large technical losses remain, tariffs do not cover cost, and allegations of corruption are widespread.

22. Governance problems remain. Governance has been improved in fiscal and monetary management, the financial sector, the exchange and trade system, and to some extent in tax administration. However, more broadly, while survey data show some improvement in governance and the control of corruption in Pakistan in recent years, the country still lags behind most of its South Asian neighbors. Corruption is widespread in public service delivery, with almost all respondents in a World Bank survey reporting corruption of one form or another. In particular, governance problems contribute to the large losses made by the power sector entities noted above, as well as to the low tax ratio.

What is still open?

23. Despite the economic recovery and increased social spending, it is as yet unclear whether poverty has declined. Comparable data to the 2000/01 household survey will become available only toward the end of 2005. However, a decline in poverty rates is expected, as average per capita incomes have risen by over 40 percent in U.S. dollar terms since 2000/01. Moreover, agricultural production has increased, though it is still unclear how well the benefits of higher growth have been distributed across the population in rural areas.

24. Some recently available key social indicators show a significant improvement over the past three years (Table 5). Survey results show a considerable improvement in a number of health and education outcomes, such as immunization, school enrollment, and literacy rates, reflecting the increase in social spending in recent years. But while social spending has increased, levels are still relatively low compared to some other countries in the region. In line with this, some social indicators remain modest compared with other low income countries (including in South Asia) (Table 6).

Table 5.

Pakistan: Key Social Indicators, 2000/01–2004/05

(In percent)

article image
Source: Pakistani authorities.

Pakistan Integrated Household Survey.

Pakistan Social and Living Standard Measurement. This survey was conducted using Core Welfare Indicators Questionnaire (CWIQ) methodology with a sample size of 76,520 household. Data are not necessarily comparable.

Table 6.

Pakistan: Social and Demographic Indicators, 2003

article image
Sources: World Bank; and UNDP 2005 Human Development Report.

III. Main Issues

A. Why the Change Between Pre- and Post-Crisis Period?

25. The weak economic performance and poor implementation of the pre-2000 programs largely reflected political instability in Pakistan, as well as the limited program ownership by eight successive governments. The main motivation for engaging the Fund tended to be short-term alleviation of foreign exchange shortages rather than sustained reforms for macroeconomic stability and long-term growth. Vested interest groups resisted deeper structural reforms while the quality of institutions eroded steadily. The top political leadership was either not committed or not able to overcome opposition to economic reforms.

26. The government that came to power in 1999 put the economy at the top of its agenda. President Musharraf took a personal interest in economic matters. Experienced technocrats, often expatriates, were recruited for key economic posts, such as the minister of finance and the central bank governor. The government did not face the same political constraints as earlier elected ones, and could take unpopular decisions, including reducing subsidies. It also dealt more efficiently with corruption at the top. Nevertheless, it still had great difficulties in expanding the tax net and getting the population to pay utility bills.

27. The initial reform efforts in 2000–01 took place against the background of relative international isolation, reflecting sanctions because of nuclear tests and opposition to the military takeover. However, with the government joining the coalition against terror in September 2001, considerable external financial support was mobilized, including more grants and a Paris Club debt rescheduling on very favorable terms. Pakistan was also compensated for costs incurred in the war on terror.

28. A frequently asked question is whether Pakistan’s improved economic performance after 1999 was the result of domestic policies or external support, and hence whether the gains are sustainable. In practice, it is impossible to precisely disentangle policy and external effects. International financial support certainly created a favorable external environment for domestic policy adjustment. It would have been impossible to secure financing assurances for the first year of the PRGF program without the change in sentiment among donors that followed the September 11 events. The agreement by the Paris Club to provide an exceptional debt treatment and the willingness of major donors to contribute fresh cash contributed to the improvement in the fiscal and external positions in 2001/02 (Tables 7 and 8). The increase in defense-related receipts provided extra resources but these were used to cover additional costs incurred in the war on terrorism (Table 9). However, without the change in policies, instigated in 1999, the international support would have fallen on infertile grounds. Instead, the increased support helped the government to advance its reform agenda and to reduce the fiscal deficit substantially. There was some early success in raising tax revenues in 2002/03–2003/04, nontax revenues (excluding defense receipts) were raised substantially, and domestically financed defense expenditures (defense spending minus defense receipts) were reduced quite sharply.

Table 7.

Pakistan: External Official Support, 1998/99–2004/05

(In percent of GDP)

article image
Sources: State Bank of Pakistan; Ministry of Finance; and Fund staff estimates.

Includes support under the Saudi Oil Facility that was phased out in 2003/04.

Includes lending by the World Bank, the IMF, and AsDB. Also includes the flow impact of debt relief which is estimated for 2003/04 and 2004/05 at 1.0 percent of GDP and 0.8 percent of GDP respectively.

Table 8.

Pakistan: Current Account Performance, 1998/99–2004/05

(In percent of GDP)

article image
Sources: State Bank of Pakistan; Ministry of Finance; and Fund staff estimates.
Table 9.

Pakistan: Fiscal Performance, 1998/99–2004/05

(In percent of GDP)

article image
Sources: Pakistani authorities; and Fund staff estimates.

29. The combination of external support and policy adjustment helped to restore confidence, resulting in lower interest rates and a reduction in the huge domestic interest bill. On the external side, the increase in private transfers, rather than official support, was the main factor behind the increase in foreign exchange reserves. This may have partly reflected tighter scrutiny of bank accounts of Pakistanis abroad following the September 11 events, but also, to a large extent, resulted from increased confidence.5 An even more pronounced surge in remittances had occurred during an earlier period of relative political and economic stability in the late 1970s and early 1980s. The considerable improvement in the balance of goods and services reflected weak domestic demand and improved competitiveness, as well as growing demand in partner countries.

B. Was Structural Conditionality Appropriate?

30. The 2002 guidelines on structural conditionality emphasized national ownership, parsimony and criticality, tailoring to circumstances, coordination with other multilateral institutions, and clarity in specification. Although the 2000 interim guidance note, which advised to limit structural conditions to macrorelevant measures in the Fund’s core areas, had not yet been finalized when Pakistan’s 2000 SBA was negotiated, the drive to streamline Fund conditionality was already well underway.

31. Conditionality in the 2000 SBA clearly did not meet the test of parsimony. In terms of numbers, structural conditions exceeded Fundwide averages by a wide margin. Moreover, conditions were not limited to just those areas that were macrorelevant and critical to the success of the program.6

32. The wide array of structural conditions in the 2000 SBA appears to have been a response to Pakistan’s weak performance under earlier Fund-supported programs. Executive Directors at the Board discussion of the 2000 SBA request in November 2000 generally considered the program to be a way for Pakistan to establish a good track record—previously it had been known as a “one tranche country.” Directors emphasized the need for prior actions and structural conditions, including at informal Board meetings prior to the approval of the program. The Acting Chairman noted that: “the uncertainty that remains has been the reason for the program to include many structural benchmarks and quantitative performance criteria.” The view that the inclusion of so much conditionality entailed unnecessary risks for program implementation was clearly not valued sufficiently.

33. In the end, the 2000 SBA was successful in establishing a good track record and in the subsequent PRGF arrangement conditionality was increasingly streamlined. But because of the large number of conditions, the impression among some Pakistani officials was that the program was punitive and imposed. In their view, the fact that it broadly succeeded had more to do with the government, with no other options available, having to accept the conditions and do whatever it took to meet them. Others, however, stressed that most, if not all, of the SBA conditions were measures the government aimed to implement in any event, regardless of the Fund. The 2001 PRGF arrangement, by contrast, was publicly claimed by the government as its own program,7 and structural conditionality was more limited and focused. It focused on three areas that were considered critical for the success of the program: (a) strengthening the tax system and expenditure management; (b) increasing the role of markets in the financial system; and (c) improving the performance of state-owned enterprises and governance, notably in the energy sector. While not a core area of Fund responsibility, the energy sector was deemed critical to achieve the program’s fiscal goals because of its large drain on budget resources.

34. But while conditionality became more parsimonious in the PRGF, it is questionable whether the chosen conditions sufficiently met the criteria of criticality, tailoring, and even ownership in some sectors. As noted above, the revenue intake did not increase as planned, and transfers to the energy sector remained large. At the same time, however, the program’s objectives in terms of reducing the debt burden and creating room for social spending were met. While tax and energy sector reforms as a whole could be considered as critical ex ante, many of the individual structural conditions were not (see below).

35. Fiscal conditionality placed emphasis on both streamlining taxation of existing taxpayers and on broadening the tax base. The authorities enthusiastically embraced efforts to make the tax system more efficient and taxpayer-friendly, including as a way to improve compliance. However, efforts to widen the tax base—by eliminating numerous exemptions and expanding sales and income tax coverage—ran into resistance from interest groups, especially as regards extending the tax net further into the agricultural and services sectors. PCs were often breached, especially the one on no new tax exemptions, but waivers were granted as the impact of the measures was minimal, which raises the question of the choice of individual measures subject to conditionality. But it is doubtful whether tougher conditionality, or implementation of conditionality, would have been successful in the absence of full ownership, including at lower levels of government, as extending the tax base into agriculture and services required cooperation from provincial authorities. A system to tax agricultural incomes, for example, pursued through Fund-supported programs since 1981, had been put into place during the 2000 SBA, but actual collection fell short of expectations, reflecting a lack of political will at the provincial level to fully implement the new laws.

36. Political constraints also prevented the government from moving faster and more decisively on power sector reforms. Electricity tariffs and bill collection remain among the most difficult political issues in many developing countries—Pakistan is not unique in this respect. Tariffs have been raised over the past decade, but not enough to cover rising costs, let alone much-needed investment to repair and expand the system. Heavy Fund conditionality in the energy sector stemmed from the fact that the World Bank’s efforts to push reforms had limited success. With the Bank stepping back at the time (see below), energy sector conditionality was included in the SBA and PRGF with a view to reducing the large drain posed by this sector on the budget.

37. With limited understanding of power sector issues, it proved difficult to select the structural conditions best suited to advance reforms. Fund conditionality focused mostly on tariff structures, financial performance of companies, and privatization. Conditions were often met, albeit mostly after delays. But slippages were never deemed serious enough, though, to warrant program interruption and waivers were frequently requested and obtained. Progress was made—including putting up KESC for sale, the break-up of WAPDA, and the design of a new tariff and subsidy regime—but technical losses in the system remained high, tariff collection weak, and as a result the sector’s overall performance did not improve much. Thus, here too, one can argue that conditionality was not sufficiently tailored to address the sector’s deep-rooted problems. Moreover, the reform of the power sector is a long-term process, with limited scope for quick results within a program’s horizon.

38. Overall, it appears that Fund structural conditionality was not well-suited for areas that require longer-term institution building and the cooperation of more than just the central government. Looking back also to the pre-1999 programs, various approaches have been tried to achieve the desired results in the tax and energy areas, ranging from developing broad sectoral plans, to large numbers of detailed measures, and to setting quantitative targets for tax collection or quasi-fiscal deficits. But without the adequate involvement of the World Bank (see below), and full government ownership at all levels, each of these approaches failed to deliver.

C. Should the Monetary Program Have Been Designed Differently?

39. Throughout much of the period under review, the SBP succeeded in keeping inflation in the low single digits. As noted above, however, in late 2003 inflation started to pick up, reaching double digits in early 2005. This raises the question whether the monetary program had become too accommodating toward the end of the PRGF arrangement.

40. The monetary program followed the standard approach of setting floors for the net international reserves (NIR) and ceilings for the net domestic assets (NDA) of the central bank. Initially, large post-September 11 inflows of capital (in particular unexpectedly high private transfers) facilitated overperformance on the NIR target, while the NDA target was met through partial sterilization of the associated liquidity growth. Monetary growth accelerated significantly in 2001/02 and eventually led to strong credit growth and a pickup in inflation, as the exchange rate was not allowed to appreciate substantially and sterilization was reversed (Figure 13). Should staff have insisted on greater exchange rate flexibility to contain inflation? Should the program instead have set ceilings for reserve money rather than NDA?

Figure 13.
Figure 13.

Pakistan: Inflation and Broad Money, 1998–2005

(y-o-y percent change)

Citation: IMF Staff Country Reports 2005, 409; 10.5089/9781451830682.002.A003

Source: Pakistani authorities.

41. The authorities initially viewed economic conditions as favorable to a relatively accommodating monetary policy stance. With the economy starting to recover, they wanted to maintain the momentum and feared that an appreciation of the nominal exchange rate, as well as an increase in interest rates, would dampen the expansion. They believed that the increase in liquidity was essentially demand driven, thereby negating the need for aggressive sterilization. Caution on exchange rate appreciation was supported by the initial expectation that the surge in both official and private inflows would prove temporary, and fears that the war on terror and the termination of the system of quotas under the multi-fiber agreement may adversely affect exports. These factors combined helped shape the authorities’ view that a nominal appreciation could result in pressures on reserves.

42. An assessment of external competitiveness fails to find evidence of a real exchange rate undervaluation. There appears to be a strong relationship between the real exchange rate and the trade balance, and the 2003–04 levels of the real exchange appear broadly consistent with a sustainable balance (Figure 14).8 The CPI-based real effective exchange rate (REER) had, on average, been depreciating by about 1 percent per year since mid-1990s in part explaining the strong growth of exports in 2001–04. While other factors have also been at work, the 2001–04 REER—unlike in the 1990s when the REER seems to have been overvalued—does not lend support to an undervaluation hypothesis.

Figure 14.
Figure 14.

Pakistan: Real Effective Exchange Rate

(year 2003=100) vs. Trade Balance (percent of GDP), 1995–2004

Citation: IMF Staff Country Reports 2005, 409; 10.5089/9781451830682.002.A003

Sources: Pakistani authorities; and Fund staff estimates.

43. Had reserve money been a performance criterion under the program, together with a NIR floor to ensure a sufficient reserve position, monetary expansion could have been contained through a more aggressive sterilization policy and higher interest rates. This would have entailed greater exchange rate flexibility with less build-up of foreign reserves—an option that would not have enjoyed much ownership in view of the authorities’ objectives to establish a strong reserve position, preserve competitiveness, and keep interest rates low. Thus, a reserve money target could have been counterproductive in the absence of any obvious initial misalignment of the real exchange rate. Rather than reflecting the architecture of the quantitative conditionality, the root of the acceleration of inflation was the emphasis placed by the authorities on growth compared with control of inflation.

44. In August/September 2003, when understandings on performance criteria for the final program year were reached, the authorities and the staff still believed that there had been a structural upward shift in money demand. This view was supported by the observation that sustained increases in money growth continued to be accompanied by little evidence of rising inflation (inflation was below 2 percent in June-August 2004). In these circumstances, fiscal policy correctly aimed at ensuring that the bulk of the official grants were either saved (i.e., the budgetary deficit excluding grants was unchanged) or allocated to development expenditures (within the spending envelope) while other expenditures were contained. The resulting fiscal consolidation was believed to keep inflationary expectations low while ensuring that the ceiling on central bank NDA would be met.

45. Since mid-2003, however, the monetary expansion was driven by domestic sources, as opposed to capital inflows earlier (Figure 15). Before, the SBP opted for accumulating foreign reserves and sterilized at least part of the liquidity effects. Starting in 2003/04, however, the SBP slowed the accumulation in reserves and cut sterilization operations. In fact, NDA became the source of monetary expansion as the central bank began financing the budget when commercial banks were no longer willing to roll-over treasury bills as interest rates were held negative in real terms. Indeed, the expansion of NDA sustained the growth in reserve money through 2004/05, thus accommodating the rising demand for private sector credit fueled by negative real interest rates. Private sector credit growth accelerated from 2.5 percent of the initial broad money stock in 2002/03 to 14.3 percent in 2003/04.

Figure 15.
Figure 15.

Pakistan: Reserve Money Growth, 2000/01–2004/05

(In percent)

Citation: IMF Staff Country Reports 2005, 409; 10.5089/9781451830682.002.A003

Source: Pakistani authorities.

46. In late 2003, staff became concerned with the steady increase in inflation, but the authorities still preferred not to tighten monetary policy to avoid choking off the growth momentum. Their view was that the increase in inflation largely reflected temporary supply side factors, especially the rise in the wheat support price. Only in 2005 did the continued increase in inflation lead the central bank to start tighten monetary policy more forcefully, although interest rates were still left negative in real terms.

47. In retrospect, with the acceleration of inflation since June 2003, the program’s NDA targets turned out insufficiently restrictive. From the program design perspective, however, inflation was below the program target of 5 percent until March 2004, while the quantitative targets for the last PRGF review were set in August/September 2003. Staff did, beginning in December 2003, consistently advise the authorities to tighten monetary policy more aggressively by raising interest rates faster. The policy dialogue recognized the inflation risk but the evidence on program ownership—and thus the expected readiness to tighten policy, if needed—was strong enough, so that the actual design of the monetary program was justified ex ante. Comfort was also provided by the program’s conservative velocity assumptions and the still very low inflatio rates (Figure 16). But as the growth of money demand started to slow, with hindsight, tighter ceilings for the central bank’s NDA could have helped to slow the rise in inflation, forcing a more rapid increase in interest rates.

Figure 16.
Figure 16.

Pakistan: Money Demand, 1999/2000–2004/05

Citation: IMF Staff Country Reports 2005, 409; 10.5089/9781451830682.002.A003

Sources: Pakistani authorities; and Fund staff estimates.

D. Did Fund-Supported Policies Help to Reduce Poverty?

48. The PRGF-supported program aimed to contribute to reducing poverty by establishing conditions for higher, sustainable economic growth, and, more directly, by creating room for higher social spending. Beyond that, the program did not include any specific measures to aid the poor. A poverty reduction strategy was developed, and if implemented is expected to result in poverty reduction over the medium and long term.

49. Who benefits from higher growth depends to a large extent on the prevailing socio-economic structure. Pakistan is a society with a small, very well-educated upper class, a still relatively small middle class, and a very large, mostly rural, lower class that is poorly educated and has little or no assets. The economic well-being of the poor largely depends on developments that are external to them, leaving them vulnerable to adverse circumstances. It will require a prolonged period of strong growth to enable low income groups to start accumulating assets, thus reducing the risk that they will fall below the poverty line if economic conditions worsen. In this regard, the past two Fund arrangements can be said to have contributed to poverty reduction, albeit over a somewhat longer horizon, by helping to establish conditions for continued strong growth of per capita incomes.

50. Considerable emphasis was placed on increasing social spending. The improvement in social outcomes could have been greater if not for the still weak administrative and institutional capacity at the local government level. At the time of the PRGF request, major efficiency gains in the delivery of key social services were expected from the devolution process. However, devolution, as well as capacity building, takes time. It would have been difficult for the Fund to place a greater emphasis on social outcomes, as this would have implied entering into areas that are well outside its areas of expertise, and where support was already forthcoming from the World Bank and the AsDB. A constraining factor has been that political devolution has yet to be fully followed up with financial and administrative devolution, to ensure that local governments have the financial means and administrative authority to carry out their increased responsibilities. The Fund could have become more involved in financial devolution issues, but it would have required taking positions on the division of resources among the provinces and districts. This is a highly complex and sensitive internal political issue, and it is unlikely that any Fund involvement would have been helpful or successful.

51. A number of specific measures may have had an adverse impact on the poor. These measures include, for example, adjustment of energy prices, reducing subsidies on or taxation of agricultural inputs, expanding the coverage of the sales tax, and the alignment of NSS instruments with market interest rates. No social impact analyses were conducted, however, and no specific consideration was given to strengthen the social safety net. The existing social safety net schemes—the Kushal Pakistan Program, the Food Support Program, and Zakat—were largely viewed as black boxes by the staff. To the extent that these programs received budget support, it was left to the government to determine whether part of the increased room for social spending was allocated to them.

52. In sum, during the period of the PRGF, growth has resumed, a poverty reduction strategy was developed, social spending was increased in line with expectations, some social indicators have improved, and it appears likely that poverty was decreased. These developments are welcome, but poverty remains widespread in Pakistan. Given the prevailing socio-economic structure, it is to be expected that government efforts, supported by the World Bank and Fund, can produce results only gradually in Pakistan.

E. Was the Size of the 2001 PRGF Arrangement Appropriate?

53. Pakistan’s 2001 PRGF was by far the largest PRGF ever in nominal terms, and one of the five largest in terms of quota (100 percent).9 In view of the quick recovery of Pakistan’s foreign exchange reserves in 2001–03 and the scarcity of PRGF resources, was such large access necessary?

54. In the fall of 2001, Pakistan faced a perilous external environment. The world economy was weakening in the wake of September 11, and war was expected in neighboring Afghanistan. Orders for exports were shrinking, and shipping costs increasing because of war risks. Export growth projections were revised downwards, resulting in a sizable balance of payments need.

55. Under these circumstances, the large size of the PRGF arrangement clearly reflected not just international political considerations but real economic needs. The briefing paper for the mission that negotiated the PRGF called for access of 60 percent of quota. The Pakistani authorities asked for higher access, and received the political support of major shareholders after the events of September 11. But the increased support was also consistent with the need to close a larger financing gap and additional uncertainty. At the Board meeting, some Executive Directors argued that access should have been even higher based on the recent strong track record of policy implementation and the strength of the program.

56. The exceptionally large PRGF loan was very effective in signaling the confidence of the international community in Pakistan, and catalyzing public and private financial flows. This support, together with political factors, encouraged Paris Club creditors to provide generous debt relief, and the World Bank and the AsDB to commit major financial support packages. Even more importantly, the signal of support from the international community helped rebuild confidence at home and encouraged large private inflows, especially remittances, which made a major contribution to the recovery of the external position.

57. In hindsight, Pakistan’s balance of payments appears overfinanced. The combination of large official and private inflows contributed to high money growth, eventually resulting in inflationary pressures. And in the end, Pakistan only drew 83 percent of the PRGF loan. Although the financing outcome was impossible to predict under the difficult circumstances of 2001, a substantially larger-than-targeted increase in official reserves could have allowed the authorities to refrain from drawings on PRGF resources or repay more expensive external debt ahead of time, thus accelerating debt reduction and entering on a faster track to debt sustainability. A few Directors pressed for (and staff discussed) Pakistan ceasing to draw on PRGF resources earlier in light of the strong foreign exchange reserves. Regarding the access, however, the balance of payments outcome was partly endogenous—smaller financial support might not have had the same confidence effect that contributed to the outcome. Also, while vulnerabilities have diminished, they still exist, which suggests that Fund support for the balance of payments was justified also on economic grounds.

F. Did Fund-Bank Collaboration Contribute to the Success of the Program?

58. There appears to have been good cooperation between World Bank and Fund staffs, especially during the crisis period of 2000–02. In the financial sector, the World Bank took the lead in supporting bank reforms (especially privatization) through a sectoral loan, key elements of which were incorporated in the PRGF; the Fund focused on reform of the NSS. In tax policy and administration, the Fund took the lead in providing strategic advice, while the World Bank has provided financial resources to modernize the tax administration. As regards poverty alleviation, Fund staff pressed for an increase of spending at the macroeconomic level and improved reporting, while the World Bank focused on building capacity in the regions, where health and education services were delivered.

59. Cooperation on energy sector issues was intense, but more difficult than in other structural areas. This largely reflected the different objectives of the two institutions. World Bank staff were intent on Pakistan developing a power system based on international best practices in the medium term and stepped away when they felt that the authorities lacked commitment. Fund staff, on the other hand, were keen to reduce electricity subsidies in the short-run and could not step away as these subsidies were a major cause of the fiscal imbalance. Although World Bank staff worked closely with Fund staff and were on board with program measures in the power sector, they tended to be skeptical of the impact of partial measures and kept a distance in view of what they perceived as the lack of ownership. This may in itself have weakened the measures. A more systematic effort to reconcile long and short-term objectives in structural reform might have helped, but ultimately it was the lack of commitment to reform by several energy sector stakeholders (outside the Ministry of Finance) that undermined reform in the power sector.

G. How Did the Work of Fund Staff on Pakistan Evolve Following the IEO Report?

60. In contrast to the IEO’s findings for the period until 1999, program design in the successor arrangements was no longer affected by pressures to overpromise and downplay risks. Staff working on Pakistan from 2000 felt no pressure to conclude programs or reviews, regardless of the possible catalytic effects, even after the events of September 11. In fact, there was considerable skepticism at the Board regarding Pakistan, especially in 2000/01. This situation was, for example, reflected in the wide array of conditionality in the 2000 SBA. The lack of political pressure on Fund staff was not recognized either in Pakistan or outside, and many observers felt that political factors were the key to Fund support.

61. The recent programs did not overpromise in terms of macroeconomic performance. The IEO had noted a tendency for overly optimistic projections for key program parameters (such as GDP growth, exports, and domestic savings and investment) and for some program targets (such as budgetary revenue) prior to 2000. But unlike the preceding period, program projections after 2000 were satisfactory. The trend in growth projections closely mirrors the trend in actual data (Figure 17).10 On average, the growth rate of GDP was overestimated by only 0.3 percentage points (compared to 1.5 percentage points in the previous programs according to the IEO). Projections for the external sector also conformed well with the time path of the actual data. The current account tended to overperform, mainly because of the abrupt and sustained increase in private transfers that was picked up by the projections only with a delay. For merchandise exports, the average forecast error was close to -4.5 percent, mostly due to conservative projections over the last two years. Fiscal balance projections under both the SBA and the PRGF were satisfactory although revenue forecasts were overly optimistic.

Figure 17.
Figure 17.

Pakistan: Were the Main Assumptions/Projections Overoptimistic?

Citation: IMF Staff Country Reports 2005, 409; 10.5089/9781451830682.002.A003

Sources: Pakistani authorities; and Fund staff estimates.

62. High expectations may have been created as regards poverty reduction and improved governance, in response to Board and management concerns for these issues. Even in the absence of specific targets, the heavy emphasis in staff reports on poverty reduction and improved governance may have suggested that major and prompt progress was possible. As noted above, some progress was made, but the challenges remain significant, and disappointment was often expressed by the Board. Staff could perhaps have emphasized more the socio-political factors constraining progress in poverty reduction and governance, and emphasized more that a longer time period would be needed to address these problems.

63. Regarding selectivity in committing Fund resources, it could be argued that the Fund should not have granted waivers for frequent breaches of structural performance criteria and interrupted the SBA and PRGF-supported programs in the view of the generally disappointing performance in the tax and energy reforms. However, macroeconomic objectives and fiscal targets were being met or exceeded, structural policies were making progress in many other areas, and overall waiver rates were lower than average in Fund programs, which appears to justify the waivers. Moreover, program interruption could have been highly disruptive, undermining the government’s reform efforts.

64. Despite the problems to which the IEO has drawn attention, without Fund (and World Bank) support during the 1990s, the economic situation at the end of that decade might well have been worse than it was. The country could have accumulated sizable external arrears or had more severe macroeconomic imbalances. By 2000, Pakistan already had a fairly liberal trade and exchange system as opposed to protectionism and exchange controls. Thus, continued Fund support probably ensured that when a more committed government came to office, it had fewer hurdles to overcome than otherwise. The counterfactual—that an economic crisis, triggering reforms, might have occurred earlier, had the Fund disengaged in the 1990s—can, of course, only be speculated at.

65. The recent programs by and large appear to have pursued the right goals. Stabilizing the macroeconomic situation was of immediate concern in 2000. The 2001 program design appropriately took a longer-term view, aiming to increase Pakistan’s growth potential, improving social outcomes, and reducing external vulnerabilities. The quantitative conditionality (as well as some prior actions) focused on consolidating macroeconomic stability, reducing vulnerability, and ensuring an increase in social spending. Unlike in the previous period, public debt sustainability was a central issue in the PRGF-supported program, and indeed was the key factor in determining the required amount of fiscal adjustment.

66. Structural conditionality focused on critical issues, but could have been more selective—individual measures subject to conditionality were often not critical. The results varied according to the authorities’ commitment in particular areas and their ability to overcome vested interests. Especially in those areas where success would have required cooperation at lower levels of government, or in public enterprises, progress appears to have been less satisfactory.

67. Article IV reports from 2000 on made a greater effort to identify risks and vulnerabilities. Issues on which the authorities and staff differed were discussed in the reports. The 2002 Article IV report was considered a good example of how to treat program review and Article IV issues in the same report, without neglecting the surveillance side. Nevertheless, to what extent the staff were able to step back and take a broader view is debatable. Alternative strategies to the PRGF were not discussed.

IV. Lessons for the Fund

68. The Pakistan experience during 2000–04 confirms that program ownership is decisive. But even in one country, it can vary widely across issues, depending on a complex set of political, economic, and security factors. The more difficult question is whether the Fund should be selective and step back, as the IEO report suggested, in cases where the political environment is not right for reform and ownership is lacking. In the case of Pakistan, it is quite possible that it would have had been more difficult for the country to take advantage of the change in government and external environment that came about in 2001 if the Fund (and other international institutions) had had disengaged earlier because of the past weak program implementation.

69. The staff experience with Pakistan underscores the complexities involved in designing structural conditionality under the Fund programs that achieves both country ownership and the objectives of the program. First, while it is relatively straightforward to identify critical issues (e.g., tax collection) it is much more difficult to find individual measures that are macrorelevant, let alone critical, to the success of a program. This is partly because many structural reforms show effects only well beyond the program horizon. Second, overall ownership of policies by a government is important, but ownership and political constraints need to be assessed also with respect to specific reforms. Without ownership at all necessary levels, it is virtually impossible to ensure effective implementation of many reforms, and there is little point in adding more conditions and trying to close every possible loophole. Third, if a measure is truly critical and there is no political will, there may be a case for stepping back, but caution, and some humility, needs to be exercised by Fund staff in determining what is truly critical.

70. Fund structural conditionality is best suited to areas where a few individual measures taken by the central government can bring progress. Where prolonged institution-building and capacity-enhancement are needed, particularly in areas outside the Fund’s core expertise, World Bank (and other IFIs) involvement is essential, and the Fund should not embark on such reforms without the World Bank fully on board. Areas that require cooperation at lower levels of government may also pose considerable difficulties, particularly when special interest groups are strong. All these considerations suggest a greater need for selectivity and leaving more areas of reform to be treated as less time-bound commitments.

71. Fund advice on tax reforms needs to balance the requirement to increase tax ratios with the need to develop an efficient, client-friendly system based on best practices. Political will and implementation capacity to bring new taxpayers into the net need to be assessed explicitly when setting conditionality. If either true commitment or technical capacity to implement revenue-gaining reforms is lacking, other reforms that enhance efficiency but are revenue-losing in the short-run may need to be delayed. With the benefit of hindsight, the reduction of tariffs and other revenue-losing measures could have been delayed until alternative sources of revenue were fully implemented. In coming years, as interest costs are likely to rise and nontax revenues may not continue to increase at the same pace as in the last few years, broadening the tax base will be crucial to meet the rising demand for social and development spending, while continuing to reduce the debt burden.

72. Pakistan’s experience with strong foreign exchange inflows raises the question of whether exchange rate adjustment should always be the preferred option to maintain price stability. This is a fairly common problem, for which there are no easy answers. Pakistan’s real exchange rate does not appear to have been significantly undervalued. A real appreciation could have made it more difficult for Pakistan’s textile manufacturers to position themselves ahead of the elimination of the quota regime. Thus, with hindsight, the SBP appears to have made the right choice when it was leaning against a nominal appreciation. However, the SBP should have moved faster in raising interest rates once inflation started to pick up, as was recommended by the staff.

73. The experience with Pakistan’s recent programs suggests that Fund staff should not overpromise in terms of short-run poverty reduction. Poverty is often caused by deeply ingrained socio-economic factors. Apart from focusing on growth in average per capita incomes, Fund staff can press for more and better targeted social spending, as well as better management of such spending programs. But these are not the only elements in a successful poverty reduction strategy, and they may not be sufficient to address poverty in the short-run.

74. Better macroeconomic governance and transparency can be achieved even in the most difficult program cases. However, overall perceptions of governance in a particular country are often determined by performance in areas such as utilities, civil service, law and order, and the judiciary, which are outside the Fund’s mandate. As the Pakistani experience shows, achieving tangible results in governance through Fund-supported programs should not be oversold.

75. Even in cases of very successful World Bank-Fund collaboration, the issue of short versus long-term reforms needs to be addressed. Rather than setting its own specific conditionality for the power sector, Fund staff could have focused more forcefully on the overall quasi-fiscal deficit of the sector and asked the World Bank staff to design the specific measures that would improve performance in the short-run, by raising the average tariff, strengthening bill collection, and reducing losses, even if there was a lack of commitment for fundamental energy sector reform. This would have required a clear accord at the senior management level.

76. The Pakistani experience shows that large, up-front Fund financial support, together with strong domestic policy implementation, can be very effective in catalyzing public and private sector support even in a very uncertain political and security situation. To consolidate success in Pakistan requires perseverance and the authorities and Fund staff need to be prepared to deal with shocks and avoid policy slippages. On the macroeconomic front, the risk of excessive inflation has become apparent and tax revenue continues to lack buoyancy. Energy sector reforms are not yet complete. Governance needs to be further strengthened in public institutions, which would also help to ensure that the benefits of growth reach the poor.

77. Pakistan and the Fund will need to continue to cooperate closely. In the context of surveillance and technical assistance, Fund staff will continue to provide advice to the authorities on the macroeconomic policy mix; financial, fiscal, and external vulnerabilities; and the creation of fiscal space to allow for increased social and development spending.

1

The team comprised Tapio Saavalainen (head), Ron van Rooden (both MCD), Taimur Baig (FAD), and Carlos Leite (PDR).

2

In addition, indicative targets were set for the net domestic assets of the banking sector (SBA only), federal tax revenues (SBA only), and social- and poverty-related expenditures. Halfway during the PRGF arrangement, the quantitative conditionality related to WAPDA was replaced by indicative targets for the accrual balance of WAPDA and the Karachi Electric Supply Corporation (KESC).

3

Fund-wide averages for GRA- and PRGF-supported programs were obtained from the report on the “Review of the 2002 Conditionality Guidelines” (refer to www.imf.org).

4

Fund-wide average waiver rates for quantitative performance criteria were 11 percent and 16 percent for GRA- and PRGF-supported programs, respectively, during 2001–03, compared to a waiver rate of 10 percent during Pakistan’s SBA and 4 percent on average during its PRGF arrangement. For structural performance criteria, the Fund-wide average waiver rates for GRA- and PRGF-supported programs during 2001–03 were 50 percent and 42 percent, respectively, compared to a waiver rate of 25 percent during Pakistan’s SBA and 45 percent during its PRGF-supported programs.

5

A part of the increase in recorded remittances represented a shift into formal channels as a result of the economic reforms.

6

Measures that can hardly be deemed macrorelevant and critical to the success of the program included, inter alia, the adoption of a plan to improve national accounts; implementation of an orderly process to resolve a dispute between two energy companies; elimination of nostro limits on banks’ foreign exchange holdings with correspondent banks; and the publication of quarterly fiscal reports.

7

The term “homegrown” had never been used before.

8

The base year for the REER index was chosen 2003 when the external trade was broadly balanced.

9

The other large access members were Democratic Republic of Congo, Honduras, Djibouti, and Sierra Leone.

10

In the accompanying figure, the first data observation on each program/review line is the point estimate for the preceding (pre-program) period. Thus, the starting point for the 2000 SBA (November 2000–September 2001) is the 1999/2000 preliminary estimate; the starting point for the 2001 PRGF (October 2001–September 2004) is the 2000/01 preliminary estimate; and the starting point for the fourth review of the PRGF (incorporating data available at November 2002) is the 2001/02 preliminary estimate. Therefore, the program projections start with the second data point in each program/review line, and a starting point, which is off the “actual” line, indicates subsequent revisions to the most recent data on which the program projections were based.

Pakistan: Staff Report for the 2005 Article IV Consultation
Author: International Monetary Fund